U.S. patent number 9,395,276 [Application Number 14/363,108] was granted by the patent office on 2016-07-19 for method and system for detection and analysis of railway bogie operational problems.
This patent grant is currently assigned to RAILWAY METRICS AND DYNAMICS SWEDEN AB. The grantee listed for this patent is Railway Metrics and Dynamics Sweden AB. Invention is credited to Helmuth Kristen, Jan Lindqvist, Jack R. Long.
United States Patent |
9,395,276 |
Kristen , et al. |
July 19, 2016 |
Method and system for detection and analysis of railway bogie
operational problems
Abstract
A method and system for detecting defects to railway wagon
wheels and to the rail. A method for detecting a wheel flat, or an
event that may cause a wheel flat to develop on a railway wagon,
including the steps of a) monitoring at least the longitudinal and
vertical acceleration of said railway wagon, and b) concluding that
a wheel flat has developed, or that there is a risk of developing a
wheel flat, if a specific acceleration pattern is monitored, said
pattern comprising a longitudinal acceleration above a first
threshold followed by a vertical acceleration above a second
threshold. A system for detecting a wheel flat of at least one
wheel of a railway wagon and the use of an acceleration sensor
mounted on the sprung part of a railway wagon for estimating the
wheel flat size of a wheel of the railway wagon.
Inventors: |
Kristen; Helmuth (Lund,
SE), Long; Jack R. (Boca Raton, FL), Lindqvist;
Jan (Stockholm, SE) |
Applicant: |
Name |
City |
State |
Country |
Type |
Railway Metrics and Dynamics Sweden AB |
Stockholm |
N/A |
SE |
|
|
Assignee: |
RAILWAY METRICS AND DYNAMICS SWEDEN
AB (Stockholm, SE)
|
Family
ID: |
47297303 |
Appl.
No.: |
14/363,108 |
Filed: |
December 7, 2012 |
PCT
Filed: |
December 07, 2012 |
PCT No.: |
PCT/EP2012/074811 |
371(c)(1),(2),(4) Date: |
June 05, 2014 |
PCT
Pub. No.: |
WO2013/083786 |
PCT
Pub. Date: |
June 13, 2013 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20150051792 A1 |
Feb 19, 2015 |
|
Foreign Application Priority Data
|
|
|
|
|
Dec 7, 2011 [EP] |
|
|
11192341 |
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01M
17/08 (20130101); B61K 9/12 (20130101); B61L
99/00 (20130101); G01M 17/10 (20130101); B61K
9/00 (20130101); G01M 17/00 (20130101) |
Current International
Class: |
B61K
9/12 (20060101); G01M 17/08 (20060101); B61L
99/00 (20060101); G01M 17/00 (20060101); G01M
17/10 (20060101); B61K 9/00 (20060101) |
Field of
Search: |
;701/19,29.1,33.4,34.2,34.4 ;73/117.01,117.03 ;246/169R,169S |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
198 26 422 |
|
Dec 1999 |
|
DE |
|
1 197 417 |
|
Apr 2002 |
|
EP |
|
5-306970 |
|
Nov 1993 |
|
JP |
|
WO 2008/141775 |
|
Nov 2008 |
|
WO |
|
Other References
International Search Report (PCT/ISA/210) mailed on Mar. 25, 2013,
by the European Patent Office as the International Searching
Authority for International Application No. PCT/EP2012/074811.
cited by applicant.
|
Primary Examiner: Frejd; Russell
Attorney, Agent or Firm: Buchanan Ingersoll & Rooney
P.C.
Claims
The invention claimed is:
1. A method for detecting a wheel flat, or an event that may cause
a wheel flat to develop, in a railway wagon, comprising the steps
of a) monitoring at least longitudinal and vertical accelerations
of said railway wagon, and b) concluding that a wheel flat has
developed, or that there is a risk of developing a wheel flat, if a
specific acceleration pattern is monitored, said pattern comprising
a longitudinal acceleration above a first threshold followed by a
vertical acceleration above a second threshold.
2. A method according to claim 1, wherein the longitudinal and
vertical accelerations of step a) are monitored on a sprung part of
said railway wagon.
3. A method according to claim 1, further comprising the step of c)
estimating a size of said wheel flat by comparing at least one
monitored acceleration of step a) with predefined criteria that
correspond to different wheel flat sizes.
4. A method according to claim 3, wherein the vertical acceleration
is compared in step c) with predefined criteria that correspond to
different wheel flat sizes.
5. A method according to claim 3, wherein step a) comprises
monitoring acceleration along three mutual orthogonal axes of said
railway wagon, wherein two of said axes correspond to the vertical
and longitudinal acceleration, respectively, and further estimating
the absolute value of an acceleration vector from the monitored
acceleration, and step c) comprises estimating the size of said
wheel flat by comparing the estimated acceleration vector of step
a) with predefined criteria that correspond to different wheel flat
sizes.
6. A method according to claim 1, further comprising using at least
one acceleration sensor mounted on a sprung part of a railway wagon
for estimating a wheel flat size of a wheel of the railway
wagon.
7. The method according to claim 6, wherein the at least one
acceleration sensor is configured to monitor acceleration along
three mutual orthogonal axes.
8. A system for detecting a wheel flat of at least one wheel of a
railway wagon, or an event that may cause a wheel flat to develop,
comprising at least one sensor for monitoring at least longitudinal
and vertical accelerations of said railway wagon, and a control
unit configured to detect a specific acceleration pattern, the
acceleration pattern comprising a longitudinal acceleration above a
first threshold, followed by a vertical acceleration above a second
threshold.
9. A system according to claim 8, wherein the at least one sensor
is configured to measure acceleration levels of up to 3.0 g.
10. A system according to claim 8, wherein the at least one sensor
has a bandwidth of about 10 Hz.
11. A system according to claim 8, wherein the at least one sensor
is configured to be mounted on a sprung part of said railway
wagon.
12. A system according to claim 11, wherein the at least one sensor
is configured to measure acceleration levels of about 0 g to 2.0
g.
13. A system according to claim 8, further comprising at least one
GPS receiver.
14. A railway wagon comprising at least one system according to
claim 8, wherein the at least one sensor of the at least one system
is mounted on a sprung part of said railway wagon.
15. A system according to claim 8, wherein the system is configured
for detection and analysis of other bogie operational defects.
16. A method for detecting a wheel flat, or an event that may cause
a wheel flat to develop, in a railway wagon, comprising the steps
of a) monitoring at least a longitudinal acceleration and a
vertical acceleration of the railway wagon, and b) determining that
a wheel flat has developed, or that there is a risk of developing a
wheel flat, when a specific acceleration pattern is detected,
wherein the specific acceleration pattern comprises a longitudinal
acceleration that exceeds a first threshold, the longitudinal
acceleration representing formation of the wheel flat, and a
vertical acceleration that exceeds a second threshold immediately
following the longitudinal acceleration, the vertical acceleration
representing a formed wheel flat.
Description
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method and system for detecting
defects to the wheels and to the rail. The invention further
relates to a system of diagnosing railway bogies problems or
defects, such as wheel flats, and the analysis and communication of
the findings.
BACKGROUND ART
Periodic unacceptable high forces generated by the wheel-rail
interaction of a moving railway wagon may be harmful to the wheels,
the bogies, the wagon and the rail. If the wagon operates with
defective bogies, including flat wheels, over a period of time, the
resultant damage may be costly in terms of wheel or rail wear or in
extreme cases result in dangerous derailments. The degree of this
danger may be much more severe if the railway wagon carries
hazardous material.
Technology has been developed to monitor performance
characteristics of a railway wagon and the rail track in order to
detect conditions that may cause damage or derailment. For example,
the truck/bogie performance has been measured by utilizing wayside
sensing, which measure forces at the wheel-rail interface. The
wayside technology may measure and detect bogies or wheel sets that
are not performing correctly, and communicate this information back
to a central office, so that the bogies or wheel sets could be
removed for service and inspection. This technology has inherent
disadvantages, however, in terms of costs and the multiple
locations across large rail systems required to identify the wagons
that are not performing safely.
A wheel flat (also known as flat spot) is a well-known defect
related to the tread shape of a railroad wheel, which decreases the
roundness of a wheel. A wheel flat may for example develop if the
wheel set of a railroad wagon is being dragged along the rail after
the wheel-axle set has stopped rotating. Faulty brakes or faulty
wheel set bearings, or other conditions that causes the wheel to
lock up while the railroad wagon is still moving, may thus cause a
wheel flat. Often, a wheel set must be replaced or the wheel set
turned if a large flat spot is detected, since a wheel flat may
cause serious derailment.
To summarize, for safety reasons, there is a need in the art for a
method of detecting a wheel-flat at an early stage so that required
actions may be performed to preclude derailment.
SUMMARY OF THE INVENTION
It is an object/aim of the present invention to provide a method
and a system for detecting wheel defects, rail defects and
derailing. It is a further object to provide a method for detecting
a wheel flat, or events that may cause a wheel flat as well as a
robust stand-alone unit that can be mounted on a railway wagon and
that detects unsafe conditions which can lead to damaging
events.
As a first aspect of the invention, there is provided a method for
detecting a wheel flat, or an event that may cause a wheel flat to
develop, in a railway wagon, comprising the steps of
a) monitoring at least the longitudinal and vertical accelerations
of the railway wagon, and
b) concluding that a wheel flat has developed, or that there is a
risk of developing a wheel flat, if a specific acceleration pattern
is monitored, the pattern comprising a longitudinal acceleration
above a first threshold followed by a vertical acceleration above a
second threshold.
The longitudinal acceleration refers to the acceleration in a
direction generally parallel with the railroad track, and the
vertical acceleration is the acceleration along an axis
substantially perpendicular to the horizontal plane, i.e.
substantially perpendicular to the longitudinal axis.
The first aspect of the invention is based on the insight that a
developed wheel flat or events that may cause a wheel flat, will
give rise to a specific acceleration pattern as seen in the
longitudinal and vertical acceleration. Thus, the inventors have
found that the development of a wheel flat is accompanied by a high
longitudinal acceleration, such as an acceleration above a specific
threshold, associated with the forming of the wheel flat,
immediately followed by a high vertical acceleration caused by the
formed wheel flat, such as a vertical acceleration above another
threshold.
The first and second thresholds may have been determined from
empirical data.
The first aspect of the invention thus provides a convenient method
for detecting wheel flats, and may thus raise an alarm before
further damage, such as wheel derailment, occurs.
It is to be understood that more than the longitudinal and vertical
acceleration may be measured. As an example, the acceleration along
three mutual orthogonal axes of the railway wagon, wherein two of
the axes correspond to the vertical and longitudinal acceleration,
respectively, may be measured in step a).
The measured acceleration may for example be the momentary
acceleration. This may thus give continual information on whether
or not a wheel flat has developed.
It is further to be understood that step a) may be performed
continually and the step of concluding that a wheel flat has
developed, or that there is a risk of developing a wheel flat, may
trigger other events, such as a more frequent detection of
acceleration or estimating the size of the wheel flat. As an
example, step b) may trigger measuring the acceleration along three
mutual orthogonal axes of the railway wagon, wherein two of the
axes correspond to the vertical and longitudinal acceleration,
respectively, and further estimating the absolute value of the
acceleration vector and estimating the size of the wheel flat based
on the absolute acceleration vector.
Consequently, the specific acceleration pattern may function as a
trigger event for more detailed monitoring of the acceleration.
In embodiments of the first aspect, the accelerations of step a)
are monitored on the sprung part of the railway wagon.
The sprung part of a railway wagon may for example be on the bogie
of the railway wagon, such as on the bogie and in the centre of the
railway wagon. The acceleration may also be measured inside a
railway wagon, such as close to or on the floor, which also is on
the sprung structure.
Consequently, the acceleration may be measured at the sprung
structure of the railway wagon, in the general area where the wagon
structure connects to an axle or a bogie. The inventors have found
that normal acceleration levels experienced by this part of the
wagon are within 0-0.5 g. Acceleration levels experienced by this
part of the wagon following a wheel derailment or a wheel flat tend
to be within 0-1.0 g, obviously more in the case of severe or
catastrophic derailment. This is further demonstrated in Example
1.3 below. Thus, the inventors have found that it is advantageous
to measure the acceleration on the sprung part of the wagon, since
it requires less rugged measurement equipment and since it
facilitates installation considerably.
The acceleration may also be measured on the unsprung parts, such
as parts attached to the wheel axle. This may provide for directly
relating the measured acceleration with the size of the wheel flat,
using input of e.g. the acceleration and the diameter of the
wheels.
In embodiments of the first aspect of the invention, the method is
further comprising the step of
c) estimating the size of the wheel flat by comparing at least one
measured acceleration of step a) with predefined criteria that
correspond to different wheel flat sizes.
Estimating the size of the wheel flat is advantageous since this
provides for different actions depending on the size of the wheel
flat. Different wheel flat sizes may require different safety
actions and it is very important that the train driver acts
according to the wheel flat size, since otherwise the railroad
track itself may be damaged. If for example the wheel flat is
severe, the train may need to stop for immediate service whereas a
minor wheel flat may only require extra service during standard
service routines or require that the train is driven below a
certain speed. In other words, there may be no need for stopping an
entire train immediately if a wheel flat has developed, given that
the train driver is aware of the wheel flat size.
In embodiments of the first aspect, the predefined criteria of step
c) are predefined acceleration levels or intervals that correspond
to different wheel flat sizes.
As an example, a low predefined g-level may correspond to a small
wheel flat and a higher predefined g-level may correspond to a
larger wheel flat. As an example, the vertical acceleration may be
compared in step c) with predefined criteria that correspond to
different wheel flat sizes.
Thus, the magnitude or pattern of the vertical acceleration may be
used for estimating the wheel flat size. Thus, step c) may comprise
comparing the vertical acceleration with predefined vertical
acceleration levels. The vertical acceleration may for example be
the momentary vertical acceleration, an integrated vertical
acceleration etc.
A step of estimating the wheel flat size by comparing the results
from step a) with predefined values means that the results from the
estimation of the acceleration may be used as a "fingerprint" that
is compared with predefined acceleration values, wherein different
predefined values or "fingerprints" correspond to different wheel
flat sizes. Thus, previous empirical data may be used to calibrate
the method with information of what results obtained from step b)
that correspond to the different wheel flat sizes. Thus, the step
of "comparing with predefined values" may include a linear or
non-linear fit of monitored acceleration data to different
calibration functions.
In embodiments of the first aspect, step a) comprises monitoring
the acceleration along three mutual orthogonal axes of the railway
wagon, wherein two of the axes correspond to the vertical and
longitudinal acceleration, respectively, and further estimating the
absolute value of the acceleration vector from the monitored
acceleration, and step c) comprises estimating the wheel flat size
by comparing the estimated acceleration vector of step a) with
predefined criteria that correspond to different wheel flat
sizes.
In step a) the acceleration may thus be measured in the direction
along three mutual orthogonal axes, i.e. in three perpendicular
directions. One direction may thus be along the vertical axis that
is generally perpendicular to the horizontal plane.
As discussed above, the acceleration vector may be determined or
measured after a trigger event, e.g. that the specific acceleration
pattern in the longitudinal and vertical acceleration is
detected.
The absolute value of the acceleration vector is defined as {square
root over ((x.sup.2+y.sup.2+x.sup.2))}, wherein x, y and z is the
acceleration along the three mutual orthogonal axes x, y and z.
This is advantageous since it provides a robust measure of the
acceleration that may be used for estimating the wheel flat size.
Measuring the acceleration in the direction along three mutual
orthogonal axes and further comparison with predefined values may
give information related to the wheel flat size.
As an example, the predefined criteria may be a specific increase
in the absolute value of the acceleration vector, such as a
specific increase that is withheld during a specific period of
time. Different increases of the absolute value of the acceleration
vector may thus correspond to different wheel flat sizes.
Measuring or estimating the acceleration vector may comprise
estimating the momentary absolute value of the acceleration vector
at the frequency at which the acceleration is monitored in step
a).
Thus, to clarify, the method of the first aspect of the invention
may comprise
a1) monitoring the acceleration along three mutual orthogonal axes
of the railway wagon, wherein two of the axes correspond to the
vertical and longitudinal acceleration, respectively,
a2) further estimating the absolute value of the acceleration
vector from the monitored acceleration of step a1),
b) concluding that a wheel flat has developed, or that there is a
risk of developing a wheel flat, if a specific acceleration pattern
is monitored, the pattern being a longitudinal acceleration above a
first threshold followed by a vertical acceleration above a second
threshold, and
c) estimating the wheel flat size by comparing the estimated
acceleration vector of step a2) with predefined criteria that
correspond to different wheel flat sizes.
As discussed above, steps a1) and a2) may be performed
continuously, or step a2), as well as step c), may be performed
after step b), i.e. step b) may function as a trigger event for
step a2) and step c).
The acceleration of step a) may for example be measured at a
frequency of 20 Hz. The absolute value of the acceleration vector
may be measured at the same frequency, and the momentary absolute
value of the acceleration vector may thus be used to estimate the
wheel flat size. The momentary acceleration may also be monitored
in order to differentiate between wheel flats of the wheels of the
railway wagon and defects of the railroad track. Defects to the
railroad track will cause a momentary change in acceleration levels
whereas a wheel flat will cause a change in acceleration during a
prolonged period of time.
Consequently, in a configuration of the first aspect of the
invention, there is provided a method for detecting defects to the
railroad track, comprising
a) monitoring at least the longitudinal and vertical accelerations
of a railway wagon travelling on the railroad track, and
b) concluding that there is a defect to the railroad track if a
specific momentary acceleration pattern is monitored, the momentary
pattern being a longitudinal acceleration above a first threshold
followed by a vertical acceleration above a second threshold.
This configuration may further comprise measuring the geographical
coordinates of the railroad wagon, such that the geographical
coordinates of the defect may be estimated, and such that it can be
determined whether the measured acceleration coincides with a
geographically fix position, perhaps detected earlier or later-on
by other systems of the same type. The geographical coordinates may
for example be measured by means of a GPS system.
Further, measuring or estimating the acceleration vector may
comprise estimating the maximum absolute value of the acceleration
vector during a specific period of time.
Thus, the momentary absolute value of the acceleration vector may
be measured during a specific time interval, and this maximum value
may be used in step c) for estimating whether a wheel flat has
developed and also the wheel flat size by comparing the maximum
value with predefined acceleration intervals or levels.
Moreover, measuring or estimating the acceleration vector may
comprise estimating the integrated absolute value of the
acceleration vector during a specific period of time.
The integrated absolute value of the acceleration vector may thus
be the sum of the momentary absolute values of the acceleration
vector measured during a specific time interval. The integrated
value may then be compared with predefined levels or intervals in
step c) to determine the size of any wheel flat.
Measuring or estimating the acceleration vector may also comprise
estimating the average absolute value of the acceleration vector
during a specific period of time.
The average absolute value of the acceleration vector may thus be
measured as the average of the momentary absolute values of the
acceleration vector measured during a specific time interval. This
average may then be compared in step c) with predefined levels or
intervals to determine the size of any wheel flat.
In the embodiments described above, the "specific period of time"
may be a long enough time interval such that any damage to the
railroad track that causes a shift in the monitored accelerometer
(a railroad track damage may cause a momentary increase in measured
acceleration) may be neglected.
The specific period of time may for example be over about 5 s.
It is also to be understood that e.g. both the momentary
acceleration as well as the integrated acceleration, an average
acceleration and/or a maximum value of the acceleration may be
monitored and estimated. This may e.g. allow for detecting both
defects to the railroad track, causing changes in the momentary
acceleration, and the wheel flat size, which may be seen as a
change in e.g. the integrated acceleration over a specific period
of time.
As an example of the above embodiments, a result from step c) of a
g measured over time interval .DELTA.t, wherein a may be the
maximum absolute value of the acceleration vector over .DELTA.t,
the integrated absolute value of the acceleration vector over
.DELTA.t or the average absolute value of the acceleration vector
over .DELTA.t, is compared with g-levels c1 and c2 (c1<c2). The
levels of c1 and c2 may be determined based on empirical data and
may be different depending on for which railway wagon the
acceleration is measured. If a is below c1, it is concluded that no
wheel flat has developed, if a is between c1 and c2, then a small
wheel flat has developed, and if a is above c2, then a large wheel
flat has developed.
As a further configuration of the first aspect of the invention,
there is provided a method of detecting events that may cause a
wheel flat, comprising the steps of
a) monitoring at least the longitudinal acceleration of the railway
wagon, and
b) concluding that there is a risk of developing a wheel flat if a
second acceleration pattern is monitored, the second acceleration
pattern being an oscillating longitudinal acceleration.
Thus, this configuration of the first aspect is based on the
inventors insight that an oscillating longitudinal acceleration
indicates that there is a risk of developing a wheel flat. Such a
scenario may for example be if the railway wagon has a locked
brake, e.g. due to malfunction or ice formation. The inventors have
found that the resulting force will cause an oscillating
longitudinal acceleration which may be monitored and used as a
warning to the train driver. This is advantageous in that it
provides for actions to be taken before an actual wheel flat has
developed.
The oscillating acceleration may be an acceleration that has a
specific magnitude in the oscillations, such as oscillations
between two predefined acceleration levels. These levels may be
determined based on empirical data.
In a similar configuration of the first aspect of the invention,
there is provided a method for detecting events that may cause a
wheel flat in a railway wagon, comprising the steps of
a) monitoring the acceleration along the direction of an axis
substantially parallel with the railroad track
b) comparing the acceleration with predefined acceleration levels,
and
c) if the monitored acceleration is above a specific acceleration
level or within a predefined interval, concluding that there is a
risk of developing a wheel flat.
In analogy with the configuration above, the inventors have found
that the above configuration may detect events that may cause a
wheel flat by monitoring the longitudinal acceleration of a railway
wagon. If the railway wagon has a locked brake, e.g. due to
malfunction or ice formation, the resulting force will cause a
longitudinal acceleration which may be monitored and used as a
warning to the train driver. This is advantageous in that it
provides for actions to be taken before an actual wheel flat has
developed.
In a further configuration of the first aspect, there is provided a
method for estimating the wheel flat size of a railway wagon,
comprising the steps of
a) monitoring the acceleration in the direction along three mutual
orthogonal axes of the railway wagon,
b) estimating the absolute value of the acceleration vector from
the monitored acceleration; and
c) estimating the wheel flat size by comparing the results from
step b) with predefined criteria that correspond to different wheel
flat sizes.
As discussed above, the acceleration may be monitored on the sprung
part of the railway wagon.
As a second aspect of the invention, there is provided a system for
detecting a wheel flat of at least one wheel of a railway wagon, or
an event that may cause a wheel flat to develop, comprising at
least one sensor for monitoring at least the longitudinal and
vertical accelerations of the railway wagon, and a control unit
adapted to detect a specific acceleration pattern, the acceleration
pattern comprising a longitudinal acceleration above a first
threshold, followed by a vertical acceleration above a second
threshold.
Terms and definitions used in connection with the second aspect of
the invention are as defined in the first aspect above.
The system of the second aspect of the invention may thus be used
in the method as defined by the first aspect above.
It is to be understood that a train may be equipped with several
systems according to the present disclosure. The systems may be
located in different railway wagons.
In the acceleration pattern, the vertical acceleration may
immediately follow the longitudinal acceleration.
The sensor for monitoring the acceleration may be an accelerometer.
The accelerometer refers to an electromechanical device that
measures acceleration forces. Such forces measured by the
accelerometer may be static, i.e. forces that do not change in
direction or amplitude, or dynamic, i.e. forces that change. The
constant force of gravity experienced on the earth's surface is
static. Forces other than gravity may be static or dynamic. For
example, vibrational movement in a rail wagon structure is
associated with dynamic forces.
Consequently, in embodiments of the second aspect, the sensor is
adapted to monitor static forces. This may determine the angle with
which the railway wagon is tilted.
In embodiments of the second aspect, the sensor is adapted to
monitor dynamic acceleration forces. By sensing dynamic
acceleration forces, one can analyse the movement of the railway
wagon.
The sensor or accelerometer may of course be adapted to measure
both static and dynamic acceleration forces.
The sensor or accelerometer further measures acceleration in three
mutual orthogonal axes, i.e. in three perpendicular directions. One
direction may be along the generally vertical axis relative to the
railroad track.
The sensor, or accelerometer, may be integrated within an
electronics module or be an externally connected accelerometer.
In embodiments of the second aspect, the sensor or accelerometer
has a digital output. A digital accelerometer tends to produce a
pulse width modulated signal: A square wave of constant frequency
may be produced, and the time interval during which the voltage is
high corresponds to the acceleration measured.
Further, in embodiments of the second aspect, the sensor or
accelerometer has an analogue output. An analogue accelerometer
produces a continuous voltage that is proportional to the
acceleration measured.
Whether to use an analogue or digital accelerometer may depend on
the hardware of the control unit with which to interface the
accelerometer. If for example a microcontroller with purely digital
input is used, a digital accelerometer is the most straightforward
solution. On the other hand, if a microcontroller with
AD-conversion capability is used, such as a PIC family one, or even
a completely analogue based circuit is used, analogue may be the
preferred choice.
In embodiments of the second aspect, the control unit comprises a
microcontroller.
The control unit may amplify and filter the acceleration signals
and also store the monitored signals or processed signal in a
storage unit.
The control unit may for example be a standard microcontroller or a
more complex microprocessor, on a printed circuit board with an
internal or external signal processor, such as BeagleBoard,
BlackFin, IGEPv2.
The control unit may be sealed in a weather proof, corrosion
resistant housing and may be connected to the sensor for measuring
acceleration or other probes on one side through a water proof
connector. This housing design allows the system to be used in an
on-wagon environment where water or moisture, dust and dirt are a
problem.
The system of the second aspect of the invention may be powered by
means of an external battery, such as a lithium-ion battery. The
system may also be powered by other means, for example by kinetic
harvesting via a hubometer or portable wind power.
In the context of the present disclosure, the wheel flat size may
be defined as the maximum decrease in radius found along the flat
part of the wheel tread surface, i.e. which is damaged by abrasion.
Such reduction of wheel radius may directly be converted into the
size or area of the actual flat portion of the wheel.
The wheel flat size may also be expressed as the length of the flat
part of the wheel tread surface, i.e. which is damaged by abrasion,
as measured parallel to the rail.
In embodiments of the second aspect, the at least one sensor is
adapted to measure acceleration levels of up to 3.0 g.
In embodiments of the second aspect, the at least one sensor is
adapted to measure acceleration levels of 0-2.0 g, such as 0-2.0 g
along three mutual orthogonal axes, wherein the vertical and
longitudinal axes are two of those axes.
One "g" is the Earth's level of gravitational force at sea surface,
i.e. 9.81 m/s.sup.2.
The inventors have realized that acceleration measurements of up to
about 3.0 g, may be sufficient to detect events such as a wheel
derailment or a wheel flat. The sensor may for example be adapted
to measure acceleration levels of 0-2.0 g along all three mutual
orthogonal axes. Thus, the system of the second aspect may be
equipped with rather non-complex accelerometers or sensors, but
still be able to give information concerning the wheel flat size.
The at least one sensor may also be adapted to measure maximum
acceleration levels of up to 3.0 g, such as about 0-2.0 g along the
three mutual orthogonal axes.
In embodiments of the second aspect, the sensitivity of the sensor
is about 0.02 g. Such sensitivity or resolution may be enough or
preferred for allowing the estimation of the wheel flat size.
In embodiments of the second aspect, the at least one sensor has a
bandwidth of about 10 Hz.
The bandwidth relates to the possible number of independent
acceleration level measurements per time unit. The inventors have
found that a bandwidth of about 10 Hz may be enough for estimating
the wheel flat size according to the present disclosure. Thus,
little bandwidth may be required for the proposed system. However,
a sensor or accelerometer having a larger bandwidth may be used.
For vehicle control or vibration measurement, a 100 Hz bandwidth or
more may be preferred. For detailed tilt sensing applications, a 50
Hz bandwidth may be enough.
In embodiments of the second aspect, the at least one sensor is
adapted to be mounted on the sprung part of the railway wagon.
This means that the sensor does not need to be mounted on the axle
or bogie of the railway wagon. The sprung part may be on the bogie
of the railway wagon, such as on the bogie and in the centre of the
railway wagon. The sprung part may also be inside a railway wagon,
such as close to or on the floor. The sensor may for example be
adapted to be mounted by magnetic means.
The sensor may be adapted to be mounted on the sprung part of the
railway wagon, close to an axle or a bogie. Normal acceleration
levels experienced by this part of the wagon are within 0-0.3 g.
Acceleration levels experienced by this part of the wagon following
a wheel derailment or a wheel flat tend to be within 0-1.0 g;
obviously more in the case of severe or catastrophic derailment.
Thus, the inventors have found that it is advantageous to mount the
sensor or accelerometer on the sprung part of the wagon, since it
requires less rugged measurement equipment and since it
considerably facilitates installation. Thus, if the sensor is
mounted on the sprung part of the railway wagon, the sensor may be
adapted to measure acceleration levels of up to 3.0 g, such as
about 0-2.0 g, such as about 0.1-1.5 g. As a further example, the
sensor may be adapted to measure acceleration levels of about a
0.1-3.0 g, such as about 0.1-2.0 g, along three mutual orthogonal
axes, wherein the vertical and longitudinal are two of those
axes.
In embodiments of the second aspect, the control unit is mounted
together with the sensor.
If for example the sensor is mounted on the sprung part of the
railway wagon, also the control unit may be mounted together with
the sensor on the sprung part of the railroad wagon. This means
that the control unit may be composed of a simple circuit instead
of e.g. a microprocessor. Further, by mounting the control unit
together with the sensor, the power consumption may be decreased,
i.e. the sensor and control unit consume little energy as long as
no problems are identified. Thus, the system may have on-board
algorithmic intelligence and may not require external data
processing for estimating the wheel flat size.
In embodiments of the second aspect, the sensor or accelerometer is
mounted directly on the axle or in the vicinity of an axle, and the
control unit is mounted on the sprung part of the railway wagon.
This could be an option in applications where more distinct
acceleration signals are required.
In embodiments of the second aspect, the system is further
comprising a wireless transceiver for transmitting acceleration
data and/or information about the wheel flat.
The wireless transceiver may be mounted together with the at least
one sensor and/or together with the control unit. If the sensor and
control unit is mounted together, the wireless transceiver may
transmit information about the wheel flat to e.g. the train driver
or a person that directs and facilitates the movement of trains
over an assigned territory, such as a rail traffic controller. If
the sensor and control unit are mounted separately, the wireless
transceiver may transmit acceleration data from the sensor to the
control unit for further analysis and further transmit information
from the control unit to e.g. the train driver. The wireless
transceiver may for example be a GPRS-unit. However, an additional
low-power local-area wireless network communication function, such
as provided by the Bluetooth or Zigbee technologies, may also be
used in order to provide communication with e.g. the train driver
in the absence of GPRS coverage.
In embodiments of the second aspect, the system is further
comprising a storage unit for storing the monitored acceleration
data.
The system may thus basically gather accelerometer data and store
the data in the storage unit, and further transmit the data via
GPRS with the transceiver to the control unit when there is GPRS
coverage. If GPRS communication cannot be established due to lack
of GPRS coverage, the storage may store the acceleration data
locally to be transmitted as soon as GPRS communication is
established. Once data is transmitted successfully via GPRS, any
data that has been stored in the storage unit may be deleted.
The storage unit may be a part of the control unit, and may be
mounted together with the at least one sensor.
In embodiments of the second aspect, the system is further
comprising at least one strain gauge.
This may give further information on other bogie operational
defects,
The strain gauge may be mounted together with the at least one
sensor
In embodiments of the second aspect, the system further comprises
at least one GPS.
A system comprising a GPS would provide measurements with
information about time and position of the railway wagon. The GPS
may be mounted together with the sensor and/or control unit.
Further, a GPS may identify that sporadic acceleration is caused by
a geographically fix spot, i.e. that it is caused by a "rail" flat
or defect, as opposed to a wheel flat.
The system may also further comprise a temperature probe and/or a
clock.
In embodiments of the second aspect, the system is furthermore
arranged to perform preliminary data processing to determine
critical and non-critical defects into three action related
characteristics: 1) Imminent derailment danger, 2) maintenance
needed before wagon can be put back in service and 3) maintenance
action needed in the future.
The system may furthermore be arranged to, after preliminary data
processing as described above, immediately send imminent derailment
danger messages to the train driver in order for him to stop the
train.
The system may furthermore be arranged to send the other two alert
levels to a back office server for additional processing. The back
office server may be arranged to prepare and send reports to
customers or provide access to the data for their account over the
internet. The back office server may furthermore be arranged to
analyse, compare and/or combine the data with stored data relating
to the same wagon, i.e. previous wagon history.
In embodiments of the back office server, it may be arranged to:
Receive and store acceleration and time stamp data, and optionally
GPS data, from each train in a structured form. Compile this
information, by background processes, into sets of data that
represent train or wagon specific information of interest, such as
maximum acceleration levels and associated geographic position.
Trigger functions that warn the user e.g. in case of acceleration
levels in excess of a given threshold, a wheel set in excess of a
given mileage or repeated acceleration peaks at a certain
geographic position.
In embodiments of the second aspect, the system functions may be
described as: Acceleration measurement in 3 dimensions at 12 Hz
intervals, with time stamp. Digital Signal Processing of the
acceleration measurements such as thresholding, averaging as well
as frequency analysis. Potential for integrating further sensor
modules, such as a GPS module or temperature sensor. Secure
wireless communication via for example GPRS. Continually storing,
by the CPU, of each accelerometer measurement to a local memory,
for example a local flash memory in the format of a RAM-disk.
Connection, by separate CPU process(es), to the GPRS network at
predetermined intervals, for example every 15 minutes. If the
connection is successful, data may be packaged and encrypted and
then transmitted to a back office server. If a receipt is received
from the back office server that the transmission has been
successful, the corresponding data is removed from the local
memory. If connection either cannot be established, or a
transmission attempt is unsuccessful, data is kept until the next
transmission attempt, i.e. for example 15 minutes later.
As a third aspect of the invention, there is provided a railway
wagon comprising at least one system according to the second
aspect, wherein the at least one system is mounted on the sprung
part of the railway wagon.
The third aspect thus provides a railway wagon in which the wheel
flat size may be monitored in convenient ways using the system as
disclosed in relation to the second aspect above.
The present invention further provides the use of a system
according to the second aspect above for detection and analysis of
other bogie operational defects, such as derailment, bogie hunting
or stuck brakes or for detection and analysis of broken rail, rail
defects or sun kinks.
As a fourth aspect of the invention, there is provided the use of
at least one acceleration sensor mounted on the sprung part of a
railway wagon for estimating the wheel flat size of a wheel of the
railway wagon.
The terms and definitions used in relation with the fourth aspect
are as defined in relation to the other aspects of the invention
above.
As discussed above, the inventors have found that it is
advantageous to measure the acceleration on the sprung part of the
wagon, since it requires less rugged measurement equipment and
facilitates installation considerably.
In embodiments of the fourth aspect, the at least one acceleration
sensor is adapted to monitor the acceleration along three mutual
orthogonal axes.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows the high-level architecture of a system according to
the present disclosure. FFT--Fast Fourier Transform, GPS--Global
Positioning System.
FIG. 2 shows the functional components of a system according to an
embodiment of the present disclosure.
FIG. 3 shows schematic drawing of a railway wagon with a sensor
mounted on the sprung part.
FIG. 4 shows a sinusoidal wheel displacement over the wheel
flat.
FIG. 5 shows example acceleration level measurements performed on a
sprung railway wagon structure.
FIGS. 6a-f show example acceleration level measurements performed
on a sprung railway wagon structure with and without wheel
flats.
FIGS. 7a-f show histogram plots of the acceleration level
measurements shown in FIGS. 6a-f.
DETAILED DESCRIPTION OF THE INVENTION AND SCENARIOS
FIG. 1 shows a high level architecture of a system according to the
present disclosure. The system may warn against defects as
derailment, wheel flats, indicate wheel flat size, warn against
hunting, damage to other rail car components, rail defects, and
severe defects in the suspension system of an individual railway
wagon.
The signal generation may comprise sensors such as at least one
accelerometer and strain gauges for monitoring the acceleration and
vibrations. The at least one accelerometer monitors acceleration in
three dimensions continually and real-time analysis of such
acceleration data identifies signatures for the defects mentioned
above. Depending on problem type and severity a warning is
communicated to the train driver or a train dispatcher. Sensors for
temperature etc. may also be included in the system if
required.
The signal generation may also include calibration signal sources,
such as from a GPS (time-velocity-position) etc. Further,
amplifiers may be included to amplify one or several of the
individual signals in the signal channels, as well as low pass
filters to purify the signals and to eliminate high frequency data
(normally above 10 Hz), which do not contain enough energy to be
damaging. Moreover, analogue-to-digital converters, ADCs, may be
included for analogue to digital conversion, which may be useful
for analysis in the digital domain.
The signal processing may include Peak and Hold circuits which
measure peak values over predetermined intervals over the complete
low pass regime. The signal processing may also comprise Fast
Fourier Transforming (FFT)-circuits that permit analysis of forces
at preselected frequencies. Further, the signal processing may
comprise correlator circuits that permit isolation and
amplification of forces that are only of significance if detected
from more than one source simultaneously. A multiplexer, MUX, may
also be required to permit sending information from a number of
inputs through a single channel.
The signal processing and application processing may be performed
in a control unit. The application processing may comprise a
micro-controller which does the preliminary data processing,
utilising proprietary algorithms, which may be required for
information to be sent directly to the train driver. It may also do
preliminary processing of data to be sent to a central server for
more sophisticated analysis using additional proprietary
algorithms.
The communication unit may include a wireless transceiver
(transmitter amplifier and antenna) for transmission of data to the
train driver or an external computer system.
FIG. 2 shows the functional components according to an embodiment
of the present disclosure.
The sensor comprises a 3D accelerometer with a lateral sensitivity
of 0.1-3.0 g, a longitudinal sensitivity of 0.1-3.0 g and a
vertical sensitivity of 0.1-3.0 g. The resolution of the
acceleration measurements is 0.02 g and the sampling interval is
about 50-100 ms.
The system further comprises a control unit comprising a
microprocessor connected to a storage unit (denoted "memories") as
well as a signal processing unit for processing the monitored
acceleration signals. A GPS is also connected to the microprocessor
to log the time and position of the monitored acceleration data.
The microprocessor analyses processed acceleration signals and
compares them with predefined values or functions in order to
decide whether or not a wheel flat has developed and the wheel flat
size. The information is sent with a transceiver, in this case a
GPRS unit (comprising a baseband processor, a, transmitter, an
amplifier and an antenna) e.g. to the train driver.
FIG. 3 shows a schematic drawing of a part of a railway wagon
comprising a bogie 1. The bogie 1 comprises unsprung bogie
structures 1a and a sprung bogie structure 1b, which is in sprung
connection 4 with the wheel pairs 2a and 2b, mounted on axles 3a
and 3b. The acceleration sensor, control unit and transceiver may
be mounted as a single package 5 by steel fitting with guides,
welded to the central part of the sprung part of the bogie 1b, or
mounted by means of strong magnets. The acceleration sensor,
control unit and transceiver may also be mounted as a single
package 5' inside the actual wagon 6, such as close to the floor of
the actual wagon 6.
The following examples further explain how the system of the
present disclosure system functions in different scenarios:
Scenario 1: Derailing
The train driver conducts a freight train along the main line. Each
wagon of the freight train is equipped with two systems of the
present disclosure, mounted close to each bogie, on the sprung part
of each wagon. After passing a junction, a wagon axle in a
monitored bogie structure derails. The increased acceleration
forces that result from the wheel derailment will exceed a
pre-defined level. This event is identified by the system. Scenario
2: Wheel Flat The train driver conducts a freight train along the
main line. Each wagon of the freight train is equipped with two
systems of the present disclosure, mounted close to each bogie, on
the sprung part of each wagon. Each system continually transfers
acceleration measurements to a central server via GPRS. In the
absence of GPRS coverage the measurements are stored in a storage
unit for later transfer to the server. A stop signal causes the
train driver to brake the train. During the brake event, one brake
in a monitored bogie structure accidentally locks and the braked
wheel loses adhesion to the track. The friction that develops
between the wheel and the track creates a wheel flat. The train
comes to a halt. After a while the line is clear and the train
driver accelerates the train. The increased acceleration forces
that result from the wheel flat will exceed a pre-defined level.
This event is identified by the system. The following scenarios
further illustrates how different operational defects may be
detected: Scenario 3: Broken Rail or Rail Defects The broken rail
signature may produce vertical acceleration levels that are
comparable to what is seen when a wheel flat occurs, or in severe
cases larger or much larger levels than what is seen when a wheel
flat occurs. However, additional geo-position information may be
used to discriminate between the two damaging events:
If a single sensor unit measures vertical acceleration levels that
repeatedly are above a certain damaging level, this may be
indicative of a wheel flat.
If a single sensor unit measures only one or a few vertical
acceleration levels above a certain damaging level, this may be
indicative of a rail defect. In particular, if other units, such as
units mounted on other trains, sense comparable levels at the same
geo-positional point at another point in time, this is a strong
indication of a rail defect. Rail defects tend to occur as a train
passes. A train that is equipped with acceleration sensors in front
and back will then be able to detect also when in time a rail
defect occurs, as the front sensors will sense no acceleration
event, whereas the back sensors will sense an acceleration event.
Scenario 4: Bearing Faults Bearing faults may increase bearing
temperature. Bearing temperature can be sensed using e.g. a single
photometric cell that derives temperature by fitting measurements
at a number of infra red frequencies to a black body spectrum.
Temperatures that lie outside a normal range, i.e. may be
indicative of a bearing fault, may be identified by either their
absolute levels, or by comparing a number of bearing temperatures
as measured on a single bogie or wagon. Alternatively, bearing
faults may be detected using acoustic measurements. Acoustic
signals originating from the bearing may be measured using a
microphone or other acoustic measurement means. The acoustic
signals may be used as an indicator for an intermediate or
non-critical warning level. It may be advantageous to receive such
an intermediate or non-critical warning level in order to
preventively replace a deteriorating bearing before sending a
railway wagon on a long trip. Acoustic signals that lie outside a
normal range may be indicative of a bearing fault, and may be
identified by either their absolute levels, or by comparing a
number of acoustic signals as measured on a single bogie or wagon.
Scenario 5: Sun Kinks A sun kink refers to bucklings in the rail
track that may occur on hot days, i.e. when the temperature of the
rail track is increased. The phenomenon may be detected by the
device, e.g. by detecting lateral accelerations above a specific
threshold in several railroad wagons in a single train, such as a
in several railroad wagons in a row. Consequently, a train may be
equipped with several systems according to the present disclosure.
Scenario 6: Hunting Hunting oscillation is an unwanted swaying
motion of a railway wagon or bogie, i.e. an unwanted lateral
oscillating movement. Such oscillation may occur if the railway
wagon or bogie travels at too high speed, i.e. above a critical
speed. If the wheels are defect in the sense that they have reduced
degree of taper or conicity and/or reduced flange thickness,
hunting oscillation may occur also at lower speeds. This event may
be identified by the system by monitoring at least the lateral
acceleration in a specific railway wagon, and conclude that a wheel
defect has occurred if a specific acceleration pattern is
monitored. The acceleration pattern may comprise a lateral
acceleration above a first threshold.
EXAMPLES
The following examples further show in detail how an acceleration
signal may be processed in order to identify and estimate defects
such as the wheel flat size. 1. Detecting Wheel Flats by Means of
On-Board Acceleration Analysis A simplistic kinematic analysis
follows that tries to identify how a wheel flat manifests itself
through acceleration. No dynamic effects are taken into account,
e.g. acceleration leading to the wheel loosing contact with the
rail. 1.1 Frequency/Duration A wheel diameter of 0.920 m (2.89 m
circumference) at a speed of 50 to 90 km/h (13.9 to 25 m/s) yields
a wheel rotation frequency of 5 to 9 Hz, corresponding to a time
period T of 200 to 111 ms. Doubtless there tend to be numerous
other wheel or axis defects that manifest themselves around these
frequencies, the first-order frequency of wheel rotation. One
possible wheel flat characteristic may be the amount of
acceleration experienced in total over one wheel rotation, another
may be the peak acceleration experienced over one wheel rotation.
Assuming a wheel flat size of 5 cm length, acceleration will mainly
occur over a fraction 0.05/2.89=1.7% of one wheel rotation, roughly
200*0.017-111 *0.017=3.4-1.9 ms. The signal amplitude needs to be
stronger than the tolerance for other first order wheel or axis
defects. It may thus be preferred that the sensor is responsive in
the interval 5-9 Hz and in the acceleration regime experienced. 1.2
Amplitude Assuming that a wheel flat size corresponds to 1 mm
abrasion/wear at the deepest spot, i.e. the diameter of the wheel
is at most reduced by 1 mm, the wheel axle is displaced by 1 mm
over time .DELTA.T=3.4-1.9 ms. FIG. 4 shows the sinusoidal wheel
axle displacement over the wheel flat
D(t)=0.460(0.ltoreq.t+t.sub.0<t.sub.0 or
t.sub.0+.DELTA.T.ltoreq.t+t.sub.0<T) [m]
D(t)=0.460-0.0005*(1-cos(2.pi.t/.DELTA.T))(t.sub.0.ltoreq.t+t.sub.0<t.-
sub.0+.DELTA.T) [m] D'(t)=0(0.ltoreq.t+t.sub.0<t.sub.0 or
t.sub.0+.DELTA.T.ltoreq.t+t.sub.0<T) [ms.sup.-1]
D'(t)=-0.0005*2.pi./.DELTA.T*sin(2.pi.t/.DELTA.T)(t.sub.0.ltoreq.t+t.sub.-
0<t.sub.0+.DELTA.T) [ms.sup.-1]
D''(t)=0(0.ltoreq.t+t.sub.0<t.sub.0 or
t.sub.0+.DELTA.T.ltoreq.t+t.sub.0<T) [ms.sup.-2]
D''(t)=-0.0005*4.pi..sup.2/.DELTA.T.sup.2*cos(2.pi.t/.DELTA.T)(t.sub.0.lt-
oreq.t+t.sub.0<t.sub.0+.DELTA.) [ms.sup.-2] Vertical
acceleration will then follow
Q(t)=D''(t)=0(0.ltoreq.t+t.sub.0<t.sub.0 or
t.sub.0+.DELTA.T.ltoreq.t+t.sub.0<T) [ms.sup.-2],
Q(t)=D''(t)=-0.0005*4.pi..sup.2/.DELTA.T.sup.2*cos(2.pi.t/.DELTA.T)(t.sub-
.0.ltoreq.t+t.sub.0<t.sub.0+.DELTA.T) [ms.sup.-2]. It is to be
noted that these levels apply to the unsprung part of the bogie.
Under these assumptions, the maximum acceleration experienced will
be 0.0005*4*3.14*3.14/(0.0034*0.0034)=1700 ms.sup.-2=170 g. Thus,
this model indicates that, should a wheel flat develop,
acceleration levels of the unsprung parts of the railway wagon may
reach 100 g, as compared to experienced accelerations of up to a
few g, such as 0-2.0 g, on the sprung parts of a railroad wagon.
1.3. Acceleration Level Estimates Acceleration level measurements
were performed on a sprung railway wagon structure. The specified
system performance with respect to acceleration measurement is
pictured in FIG. 5. The accumulated percentage of measurements is
plotted against the absolute acceleration force measured. The
measurement unit is g. As seen in the plots, 100% of the
measurements give 0 g or more, few measurements give 0.5 g or more.
The example illustrates the difference in acceleration levels
between welded track (left), that simulates a situation without a
wheel flat, and jointed track (right), which simulates the
situation with a wheel flat. Consequently, the acceleration on the
sprung parts of a railway wagon tends to fluctuate up to roughly
1.0 g, as compared to the 30-100 g that is experienced directly at
the wheel against the railroad track (see 1.2 above).
Numerous comparative tests of acceleration levels have been
conducted in-yard as well as on the main line. Acceleration levels
were gathered at various speeds, from wagons of similar type, some
with pronounced wheel flats, loaded as well as unloaded. The
measurement equipment was mounted on the sprung part of the
respective wagon, close to one of the wagon bogies. Acceleration
data was measured at 12 Hz intervals in three dimensions, and
transmitted continuously every 15 minutes via GPRS to a back-office
server function. Test time durations were up to several hours. The
results are shown in FIGS. 6a-f and 7a-f. FIGS. 6a, 6c, and 6e show
time resolved lateral, longitudinal and vertical acceleration level
measurements, respectively, on wagons with wheel flats. FIGS. 7a,
7c and 7e show corresponding histogram plots. FIGS. 6b, 6d, and 6f
show time resolved lateral, longitudinal and vertical acceleration
level measurements, respectively, on wagons without wheel flats.
FIGS. 7b, 7d and 7f show corresponding histogram plots. A clear
increase in vertical acceleration levels is measured in the case of
a wheel flat (compare FIGS. 6e and 7e with FIGS. 6f and 7f). It can
be concluded that an increase in z-level acceleration levels, as
measured on the sprung part of the wagon, is indicative of a wheel
flat.
Although exemplary embodiments of the present invention have been
shown and described, it will be apparent to the person skilled in
the art that a number of changes and modifications, or alterations
of the invention as described herein may be made. For example, the
monitored accelerations may be used to conclude that defects,
similar to but other than explicitly disclosed above, have occurred
by detecting a specific acceleration pattern. Furthermore, the same
acceleration sensors may be used to detect several types of
defects, as described above. Alternatively, a plurality of systems
may be used on each wagon for detection of several defects. It is
to be understood that the above description of the invention and
the accompanying drawings are to be regarded as non-limiting
examples thereof and that the scope of the invention is defined in
the appended patent claims.
* * * * *